From the paper: "To provide context, we also model what would happen if we did not open Cornell for a residential fall semester and did full virtual instruction

From the paper: "To provide context, we also model what would happen if we did not open Cornell for a residential fall semester and did full virtual instruction instead. Our nominal parameters assume that 9000 students would remain in Ithaca but outside the control of the University in off-campus apartments" I didn't dig deep enough to see how this assumption affected their "virtual instruction" danger rating, but it seems weird to me to assume that 9000 students would move back to Ithaca to live off campus for a fully online semester?